------------------------------
On Fri, Jun 27, 2014 10:29 PM BST ahi wrote:

>I have added some content to the wiki
>https://github.com/ah450/sdpx-tools-python/wiki , will add more
>tomorrow.

You should mention that in the readme also. Any amount is better than none...

>let me know if I am going in the right direction with the content of
>the documentation.
>I have updated the README with an example and added two example files
>one for parsing a tag/value file and one for writing a tag/value file.
>
>I am currently working on the following tasks:
>
>Updating the wiki : continuing to write documentation for already
>implemented featues, and documenting new features before working on
>them. Implementing package, license and file model validations. Adding
>extracted license rules for tag/value parser.

I have my reservations about "already implemented features", seeing
as there had been instances of codes being written which aren't used/tested.

I think at some point you should look into coverage: what part
of code are being used by the top level nosetests, and what
are not; and either hook them up for testing also, or document
which code paths are not exercised regularly. Look very carefully
at each of them for mistakes, or whether you actually need them at all,
and may better just remove them.

If it were C code coverage, I would go about it via running things
under a profiler and compare the lists of symbols of the
binary/library from nm (of GNU binutils) versus the profiler
statistics - routines in a binary not showing up in profiler stats are obviously
unused. There might be a similiar trick in python with its
built-in profiling facility.

I see you have finally got rid of all the persistent warnings that has been
going on for weeks. Congratulations. Please try not to introduce new
errors/warnings from now on.


_______________________________________________
Spdx-tech mailing list
[email protected]
https://lists.spdx.org/mailman/listinfo/spdx-tech

Reply via email to